Trade-offs and Noise Tolerance in Signal Detection by Genetic Circuits

Raúl Guantes,Javier Estrada,Juan F Poyatos

doi:10.1371/journal.pone.0012314

Raúl Guantes, Javier Estrada + Show 1 more

Open Access

https://doi.org/10.1371/journal.pone.0012314

Copy DOI

Abstract

Genetic circuits can implement elaborated tasks of amplitude or frequency signal detection. What type of constraints could circuits experience in the performance of these tasks, and how are they affected by molecular noise? Here, we consider a simple detection process–a signal acting on a two-component module–to analyze these issues. We show that the presence of a feedback interaction in the detection module imposes a trade-off on amplitude and frequency detection, whose intensity depends on feedback strength. A direct interaction between the signal and the output species, in a type of feed-forward loop architecture, greatly modifies these trade-offs. Indeed, we observe that coherent feed-forward loops can act simultaneously as good frequency and amplitude noise-tolerant detectors. Alternatively, incoherent feed-forward loop structures can work as high-pass filters improving high frequency detection, and reaching noise tolerance by means of noise filtering. Analysis of experimental data from several specific coherent and incoherent feed-forward loops shows that these properties can be realized in a natural context. Overall, our results emphasize the limits imposed by circuit structure on its characteristic stimulus response, the functional plasticity of coherent feed-forward loops, and the seemingly paradoxical advantage of improving signal detection with noisy circuit components.

Highlights

Signal transduction networks are commonly constituted by genetic circuits, or modules, comprised of a small number of interacting molecular elements
All kind of interactions are allowed between sensor and output species, and the input signal can act on both components, but there is no feedback to the input [18]
To make this definition more quantitative, we employed two detection scores: We used the output susceptibility, sO, to estimate the potential of the module to detect amplitude variation (Figure 1B and Figure S1A) [16,17,18]. This measure quantifies the relative change in the output species at equilibrium, nO, as the input signal changes –the larger sO, the better the detection, and depends on the pairwise susceptibilities between module components (Figure 1A) as sO~

Summary

Introduction

Signal transduction networks are commonly constituted by genetic circuits, or modules, comprised of a small number of interacting molecular elements. Recent accounts of how these modules compute biochemical information highlighted the intricate relationship between structure and function, and the capacity of these units to process different signal attributes (e.g., signal amplitude or frequency) [1,2,3,4,5,6]. This capacity is relevant for cellular action. The frequency of the TNFa oscillation is read by the nuclear factor kB signalling module, with different frequencies resulting in changes in timing and specificity in the transcriptional activation of downstream genes [6]

Methods

Results

Conclusion